Record variables are references to records, not the records themselves. This means that there are two different ways in which you might want to compare two record variables:
To discover if two different record variables reference the same record, you use the reference equality operator, ==
:
declare type point field integer x field integer y process local point alpha local point omega set omega to alpha do when omega == alpha output "Alpha and omega refer to the same record.%n" else output "Alpha and omega refer to different records.%n" done
To discover if two records have the same values you can do one of the following:
The following example compares two records by comparing the field values directly:
set alpha:x to 2 set alpha:y to 4 set omega:x to 3 set omega:y to 9 do when omega:x = alpha:x & omega:y = alpha:y output "The two records describe the same point.%n" else output "The two records describe different points.%n" done
The next example defines a function to compare points:
define switch function same-point (value point a, value point b ) as return a:x = b:x & a:y = b:y process ... do when same-point (alpha, omega) output "The two records describe the same point.%n" else output "The two records describe different points.%n" done
The function name "same-point" is somewhat ugly. What we really want is to use the "=" sign to test whether two points are the same. We can do this by overloading the "=" operator:
define overloaded switch infix-function value point a = value point b as return a:x = b:x & a:y = b:y process ... do when alpha = omega output "The two records describe the same point.%n" else output "The two records describe different points.%n" done
Note that there are several possible meaning that could be associated with the idea that two records are "equal" to each other. The example above uses the most obvious one, that all the fields contain the same values. However, other interpretations of the idea of equality are possible. For example, consider a record type for a place, which contains the name of the place and its latitude and longitude:
import "omfloat.xmd" unprefixed declare record place field float latitude field float longitude field stream name
Suppose we have two place records:
process local place ottawa local place capital set ottawa:latitude to 45.417 set ottawa:longitude to -75.700 set ottawa:name to "Ottawa, Ontario" set capital:latitude to 45.417 set capital:longitude to -75.700 set capital:name to "the Capital of Canada"
What is the correct way to compare these two records? A straight field by field comparison will say that they are not equal, since the name fields do not match. However, the latitude and longitude values are the same, and the two different name values are actually two different names for the same place.
In comparing two place records, therefore, we may want to compare the latitude and longitude values and ignore differences in the name field. This may be a very useful function, because it will enable us to discover if two place names are actually the name of the same place:
define overloaded switch infix-function value place a = value place b as return a:latitude = b:latitude & a:longitude = b:longitude process ... output ottawa:name || (ottawa = capital -> " is " | " is not ") || capital:name || "%n"